Atik: A Shared-Mesh Transformer Accelerator for RISC-V
By Ahmed Zeer, Erkan Uslu
Yildiz Technical University, Istanbul, Turkiye
Abstract
Transformer inference repeatedly invokes dense matrix multiplication and attention, including its data-dependent softmax normalization. We present Atik, an open-source RoCC accelerator for RocketChip that executes matrix multiplication and dense attention through one shared fixed-point MAC mesh. Atik accepts and returns BF16 tensors, uses a streaming online-softmax schedule that never materializes the full score matrix, and is invoked through a small descriptor-based ABI. We evaluate three meshes cycle-by-cycle on FireSim and report standalone Sky130 physical-design area. Across 137 logged FireSim test records, all tests pass their benchmark tolerance with zero reported output mismatches. Relative to the same scalar Rocket BF16 reference, the 8x8 configuration reaches 31.5x peak matmul speedup and 225.8x peak attention speedup. Hybrid projections for the single-layer workload-derived C models, using directly measured accelerator-kernel cycles, give aggregate speedups of 44.2x, 77.1x, and 32.3x for ViT, TinyBERT, and GPT-2 prefill, respectively; these are not full-model hardware replays.
Keywords: Transformer accelerator, attention, softmax, systolic array, RISC-V, RoCC, RocketChip, BF16, FPGA, FireSim, open-source hardware
To read the full article, click here
Related Semiconductor IP
- RISC-V Debug & Trace IP
- RISC-V IOPMP IP
- Gen#2 of 64-bit RISC-V core with out-of-order pipeline based complex
- 64-bit RISC-V core with in-order single issue pipeline. Tiny Linux-capable processor for IoT applications.
- Multi-core capable RISC-V processor with vector extensions
Related Articles
- RISC-V Based TinyML Accelerator for Depthwise Separable Convolutions in Edge AI
- A Framework for Accelerating Transformer Inference on RISC-V for Edge AI
- FlexViT: A Flexible FPGA-based Accelerator for Edge Vision Transformers
- MeshKV: A Network-on-Chip KV Cache Fabric for Scalable Transformer Decoding Accelerators
Latest Articles
- Atik: A Shared-Mesh Transformer Accelerator for RISC-V
- GSPIM: A Temporal-Locality-Aware 4D Gaussian Splatting Accelerator with Polymorphic PIM and Multi-Frame Rendering Dataflow
- Memory Admission and Result-Retention Sizing in an RV32IMC FPGA Soft Processor
- A Cross-Domain Matrix Engine for Signal Processing and Deep Neural Net Inference
- LLM Agent-Based Design of RISC-V CPUs with FPGA-in-the-Loop Evaluation